1. Volume 16, Issue 3, Pages 731-743 (July 2016)
A Role for Timp3 in Microbiota-Driven Hepatic Steatosis and Metabolic Dysfunction 
Maria Mavilio, Valentina Marchetti, Marta Fabrizi, Robert Stöhr, Arianna Marino, Viviana Casagrande, Loredana Fiorentino, Marina Cardellini, Ben Kappel, Ivan Monteleone, Celine Garret, Alessandro Mauriello, Giovanni Monteleone, Alessio Farcomeni, Remy Burcelin, Rossella Menghini, Massimo Federici 
Cell Reports 
Volume 16, Issue 3, Pages (July 2016)
DOI: /j.celrep
Copyright © 2016 The Author(s) Terms and Conditions

2. Cell Reports 2016 16, 731-743DOI: (10.1016/j.celrep.2016.06.027)
Copyright © 2016 The Author(s) Terms and Conditions

3. Figure 1 Metabolic Phenotyping of Timp3−/− Mice
(A) Fasting blood glucose levels were measured at different time points during an HFD (n = 5/each group; Student’s t test with Welch’s correction; ∗∗p < 0.01).
(B) Area under curve (AUC) during 120-min IPGTT of WT and Timp3−/− mice after 4 months on an HFD (n = 35/each group; data are mean ± SD, Student’s t test with Welch’s correction ∗∗∗∗p < ).
(C) Flow cytometry analysis of blood leukocytes was performed using monoclonal antibodies directed against monocyte markers CD11c, F4/80, and CX3CR1 (n = 5 in each group; data are mean ± SEM; Student’s t test with Welch’s correction; ∗p < 0.05, ∗∗p < 0.01).
(D) mRNA expression of CD11c, F4/80, and CX3CR1 in white adipose tissue (WAT), liver, and muscle in WT and Timp3−/− animals at 4 months on an HFD. (n = 4 in each group; data are mean ± SEM; Student’s t test with Welch’s correction; ∗p < 0.05).
See also Figure S1.
Cell Reports  , DOI: ( /j.celrep )
Copyright © 2016 The Author(s) Terms and Conditions

4. Figure 2 Inflammatory Phenotyping of Timp3−/− Mice
(A) Analysis of inflammatory infiltrate in the stromal vascular fraction (SVF) of WT and Timp3−/− animals after 4 months on an HFD (n = 5 in each group; data are mean ± SEM; ∗p < 0.05, ∗∗∗p < 0.001, ∗∗∗∗p < with one-way ANOVA with Sidak’s multiple comparison test).
(B) Expression of pro- and anti-inflammatory genes in SVF from Timp3−/− mice and WT mice (n = 5 in each group; data are mean ± SEM; ∗p < 0.05 with Mann-Whitney t test).
See also Figure S2.
Cell Reports  , DOI: ( /j.celrep )
Copyright © 2016 The Author(s) Terms and Conditions

5. Figure 3 Timp3−/− Mice Fed an HFD Have Altered BCAA and AAA Pathways
(A) Top-ten metabolites significantly different in Timp3−/− mice compared with WT mice in fasting state. Two-way ANOVA analysis adjusted for diet, with an FDR of 1%.
(B) PLS-DA (up) and random forest (down) approaches allowed to identify metabolites significantly different in Timp3−/− mice compared with WT mice on an ND or on an HFD in the fasting state.
(C) Two principal metabolites belonging to the BCAA and AAA pathways: alpha-hydroxyisocaproate and indolelactate are shown as box plots, with median, upper, and lower quartiles and maximum and mininum values in WT and Timp3−/− mice fed both an ND and an HFD in fasting state (n = 6 per group; one-way ANOVA with Sidak’s multiple comparison test; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001∗∗∗∗p < ).
See also Figures S3 and S5 and Tables S1, S2, and S3.
Cell Reports  , DOI: ( /j.celrep )
Copyright © 2016 The Author(s) Terms and Conditions

6. Figure 4
Timp3−/− Mice Fed an HFD Show Altered Expression for Metabolic and Inflammatory Signals in the Colon without Endotoxemia
(A) Real-time PCR analysis of RNA extracted from liver, muscle, white adipose tissue (WAT), and colon tissues of the Timp3−/− mice compared with the WT littermates fed an HFD. (n = 5 in liver, muscle, and WAT; n = 9 in colon; data are mean ± SEM; ∗∗p < 0.01, ∗∗∗∗p < by Mann-Whitney t test; a.u.)
(B) H&E-stained colonic sections of WT and Timp3−/− mice at 4 months on an HFD. Upper, 4×; lower, 20×.
(C and D) Real-time PCR analysis on RNA extracted from colon tissue of the Timp3−/− mice compared with the WT littermates fed an HFD. (C) Inflammatory genes, (D) wall permeability genes. (n = 9 data are mean ± SEM; ∗p < 0.05, ∗∗p < 0.01 by Mann-Whitney t test).
(E) Quantification of 16S DNA concentrations in MAT (mesenteric adipose tissue), EAT (epididimal adipose tissue), and peripheral blood mononuclear cells (PBMCs) of Timp3−/− mice compared with WT animals. (n = 5 per group; data are mean ± SEM.) Serum levels of lipopolysaccharide (LPS) in WT and Timp3−/− mice fed an HFD (n = 10 per group).
See also Figure S4.
Cell Reports  , DOI: ( /j.celrep )
Copyright © 2016 The Author(s) Terms and Conditions

7. Figure 5
Antibiotic Treatment Improves Features of Metabolic Syndrome, Low-Grade Inflammation, and ADAM17 Activity in Timp3−/− Mice Fed an HFD
(A) Fasting and fed blood glucose levels (left) (n = 6 per group; data are mean ± SEM; Student’s t test with Welch’s correction ∗∗p < 0.01) and fasting insulin level (right) in antibiotic-treated WT and Timp3−/− mice fed an HFD compared to untreated controls (n = 6 per group; data are mean ± SEM; one-way ANOVA with Tukey’s multiple comparison test; ∗p < 0.05 per fasting insulin.
(B) IPGTT test of WT (left) and Timp3−/− mice (right) after in vivo antibiotic treatment compared to an untreated control.
(C and D) Histology of liver in WT (left) and Timp3−/− mice fed an HFD (right) (H&E stain; scale bar, μm) after antibiotic treatment compared with untreated controls (C) and the percentage of microvescicular, macrovescicular, and lobular inflammation (D) (n = 4–5 per group; data are mean ± SEM; one-way ANOVA with Sidak’s multiple comparison test; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001).
(E) ADAM17 activity in untreated WT and Timp3−/− mice fed an HFD compared with antibiotic-treated littermates (n = 5 per group; data are mean ± SEM; one-way ANOVA with Sidak’s multiple comparisons test; ∗p < 0.05, ∗∗∗p < 0.001).
(F and G) mRNA expression of BCKDHα and BCKDHβ subunits in liver (F) and colon (G) before and after antibiotic treatment (Abx) in WT and Timp3−/− mice fed an HFD (n = 9 for colon analysis; n = 5 for liver analysis; data are mean ± SEM; ∗p < 0.05 ∗∗p < 0.01, ∗∗∗∗p < by Mann-Whitney t test).
See also Figure S6 and Tables S4, S5, S6, S7, and S8.
Cell Reports  , DOI: ( /j.celrep )
Copyright © 2016 The Author(s) Terms and Conditions

8. Figure 6 IL-6 and sIL6R Levels in Blood, Liver, and WAT
(A) Serum levels of IL-6 and sIL6Rα analyzed by western blot (left) and graphic representation of optical density (OD) of IL-6 and sIL6R protein levels in serum, which were normalized to loading (right) (n = 4 per group; data are mean ± SEM; one-way ANOVA with Sidak’s multiple comparison test; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001).
(B) IL-6 expression analyzed by western blot in liver and WAT of WT and Timp3−/− mice fed an HFD (left) and graphic representation of optical density (OD) (n = 4 data are mean ± SEM; ∗p < 0.05 by Mann-Whitney t test.
(C) Serum levels of IL-6 and sIL6Rα protein levels in untreated WT and Timp3−/− mice fed an HFD compared with antibiotic-treated littermates analyzed by western blot (left) and graphic representation of optical density (OD), which were normalized to loading (right) (n = 3 per group; data are mean ± SEM; one-way ANOVA with Sidak’s multiple comparison test ∗p < 0.05; ∗∗p < 0.01).
(D) ELISA of sIL6R in sera of untreated WT and Timp3−/− mice fed an HFD compared with antibiotic-treated littermates. (n = 4 per group; data are mean ± SEM; one-way ANOVA with Sidak’s multiple comparison test; ∗p < 0.05, ∗∗∗p < 0.001).
(E) Flow cytometry analysis of blood leukocyte and stromal vascular cells (SVCs) was performed using monoclonal antibodies directed against dendritic cells (CD11c+). In antibiotic-treated Timp3−/− mice fed an HFD we found a decrease in the percentage of CD11c+ population compared with littermate controls, whereas no difference was found in treated WT mice fed an HFD compared with the control after 4 weeks of antibiotic treatment (Abx) (n = 5 per group; data are mean ± SEM. ∗p < 0.05 by Mann-Whitney t test, a.u.).
Cell Reports  , DOI: ( /j.celrep )
Copyright © 2016 The Author(s) Terms and Conditions

9. Figure 7
Inhibition of sIL6R Signaling Controls CD11c+ Trafficking with Mild Effects on Glucose Tolerance and No Effect on Steatosis and ADAM17 Activity
(A) Soluble gp130Fc (sgp130Fc) effect on metabolic status: fasting blood glucose 1 month post-injection in WT and Timp3−/− mice fed an HFD injected with sgp130Fc (chimera) compared with littermates injected with PBS (vehicle) (n = 4 data are mean ± SEM; one-way ANOVA with Sidak’s multiple comparison test; ∗p < 0.05, ∗∗p < 0.01.
(B) Intraperitoneal glucose tolerance test and area under curve (AUC) (left) and fasting insulin level (right) after two months from in vivo treatment of Timp3−/− mice with sgp130Fc (chimera) compared with Timp3−/− mice fed an HFD and injected with PBS (vehicle). (n = 9; data are mean ± SEM; ∗p < 0.05 by Mann-Whitney t test; a.u. per AUC; one-way ANOVA with Sidak’s multiple comparison test per fasting insulin; ∗p < 0.05, ∗∗p < 0.01).
(C–F) Effect on CD11c+ cells in blood at 1 month (C) and 2 months (D) and in SVC at 2 months (E) after injection of soluble gp130Fc in Timp3−/− mice and in SVC at 2 months (F) after injection of soluble gp130Fc in WT mice (n = 4/each group; Bonferroni’s test ∗p < 0.05).
(G and H) Histology of liver in WT and Timp3−/− mice fed an HFD (H&E stain: scale bar, μm) after treatment compared with littermate controls (G) and the percentage of microvescicular, macrovescicular, and lobular inflammation (H).
(I) ADAM17 activity in WT and Timp3−/− mice fed an HFD and injected with sgp130Fc (chimera) compared with littermates injected with PBS (vehicle) (n = 4 data are mean ± SEM; one-way ANOVA with Sidak’s multiple comparison test; ∗∗∗p < 0.001).
See also Figure S7.
Cell Reports  , DOI: ( /j.celrep )
Copyright © 2016 The Author(s) Terms and Conditions